Generally, a colour image is represented by the three (3) primary colours of light (Red, Green and Blue). But if you represent an image by the three (3) primary colours of light, you have the large amount of data on the image because each of the three (3) primary colours has strong correlation to the other and has redundant data.
So, there are a lot of conventional methods for representing colour images and one of the conventional methods is that representing colour images by luminance, hue and saturation that can reduce redundant data between the colour image factors. These values representing the colour image are being used in compressing, extracting the characteristic value of colour images, etc.
But, even though the conventional method is an efficient way of reducing the redundant data between the three (3) factors of colour image, it complicates mathematical computation and image process in compressing and extracting the characteristic value of colour image because it still uses the three (3) factors of colour image.